Thank you very much for this great questions.

I think there is no problem if you use the same threshold on health care for all diseases and health related conditions or specific illness.

Be careful about thresholds in-terms of denominators. To date researchers have failed to accept a unique threshold for defining catastrophic expenditure. However, the most common thresholds for household total consumption expenditure or income is more than 15%, for non-food expenditure is more than 25% or 40% and household capacity to pay is more than or equal to 40%.

In addition to what Mizanur Rahman suggested you may use following references for methodology purposes.

Xu K, Evans DB, Kawabata K, Zeramdini R, Klavus J, Murray CJL: Household catastrophic health expenditure: a multicountry analysis. Lancet 2003, 362: 111-117

Xu K: Distribution of Health Payments and Catastrophic Expenditures Methodology. Geneva: Department of Health System Financing, WHO; 2005.

Rahman MM, Gilmour S, Saito E, Sultana P, Shibuya K: Health related financial catastrophe, inequality and chronic illness in Bangladesh. PLoS ONE 2013, 8: e56783.

Xu et al. references is excellent for estimating catastrophic expenditure. In addition, you can follow another paper for detail information.

O'Donnell O, Van Doorsslaer E, Wagstaff A, Lindelöw M: Analyzing health equity using household survey data: a guide to techniques and their implementation, vol. 434: World Bank Publications; 2008.

This paper provide theoretical, practical, and statistical package information for health financing assessment.

Household catastrophic expenditures are different from catastrophic illness expenditures. The first will be dependent on household income distribution and social health protection, whereas the latter could be categorised by disease diagnosis and coverage with payment limits on the supply side. Should catastrophic be defined by disease severity and related treatment costs, or by affordability of acceptable levels of care?

Dear Kai Hong Phua,

We are planning doing a study on tobacco related diseases and trying to estimate financial burden and detect presence if any, catastrophic expenditure due to that single episode of hospitalization. How should we go about it? We are planning to calculate total direct, indirect costs, estimate opportunity costs as well expenditure on tobacco and per capita income along with non-food expenditure. What else should we do that will make more accurate estimation of catastrophic expenditure?

Kindly give suggestions...

I think for your study purposes you have listed all elements that will help you estimate was or not the expenditure related to a case of hospitalization catastrophic to a household (or to an individual) or not. Please keep in mind what would be your unit of analysis a) household that had member that was hospitalized or b) an individual. Depending on choice you either need household level data or individual level data. Also be aware that if you only look at hospital expenditure, you are missing other health spending that could have happened within a household or for an individual. Also to compare your data with others make sure your hospital expenditure is annualized (for this you may need to impute other cases of hospitalization that may have happened in a household over 12 month period and their estimated cost) and related to annual income/spending. Although on a latter point you could find (though limited) other literature that has only looked at a case of medical treatment resulting in a catastrophe. Hope this helps with your study.

Daivadanam et al. have published an article in NCD area. Please visit

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3516025/

Dear Reshu,

In my experience catastrophic healthcare expenditures strongly depend on the specific health related conditions. For example, if you have large series of data available, Clinical Risk Groups (CRG) is a risk adjustment system focused on acute, chronic conditions, multimorbidity and catastrophic conditions. CRG requires individual information on demographic characteristics, care episodes (diagnostics and procedures, including primary care) and pharmacy. Catastrophic conditions in the CRG system are: 9010 Dialysis with Diabetes, 9020 Dialysis without Diabetes, 9030 HIV Disease, 9040 Total Parenteral Nutrition, 9050 Dependence on a Mechanical Ventilator, 9060 History of a Major Organ Transplant, 9070 Congenital Quadriplegia, Diplegia or Hemiplegia, 9080 Acquired Quadriplegia or Permanent Vegetative State, 9090 Spina Bifida, 9100 Progressive Muscular Dystrophy or Spinal Muscle Atrophy and 9110 Cystic Fibrosis. These categories include different groups of patients incurring different (but always high) levels of expenditures. Hope this helps you.

Cost benefit analysis

In typical chronic deseases economical evaluation (Markov Model) catastrophic expenditure are depend on the specific health related conditions/status. Framingham risk score for hearth is higher for the most common chronic diseases , so catastrophic expenditure are different for chronic and non chronic diseases .

As far as I understand, the principle of catastrophic health expenditure is the same for health in general and for NCDs in particular, to figure out whether or not the household incur financial hardship due to OOP for healthcare. However, the methodology of Xu et al couldn't be a good method in your study because it needs to have a nationwide database. In your study, I suggest you should use the indicator OOP for NCDs/ total consumption, with cut-off point 10%, 15%, 25%. (especial 10%, because it have been seen in literature.)

There would be a lot of NCDs: CVD, cancers, DM, ... So in case you would like to estimate the tobacco related diseases, choose the NCDs suitable with your purpose and focus on them (It's interesting to have pictures of catastrophic of different specific kind of NCDs....)

http://www.funsalud.org.mx/competitividad/financiamiento/Documentos/IDRC/IDRC-2DA%20EDICION.pdf

Dear Dr. Le Hong:

Thanks for your thoughtful explanation.

I am bit worried to use total consumption as a denominator only. It may not be reflected true incidence of catastrophic expenditure in low-and middle-income countries. More explanation about it in "Analyzing Health Equity Using Household Survey Data, World Bank" paper.

It may better to use different definition to assess financial burden.

For example,

Total consumption: 10%

Non-food consumption: 25, or 40%

Capacity to pay: >=40%

Leave it to readers for judgments.

Thank you again for your nice comments.

Dear Dr. Mizanur Rahman,

Thank you for your sharing. I totally agree with you it's better to use different definitions to calculate catastrophic incidence.

However, I'm a little bit concerned about the cut-off point of catastrophic using capacity to pay (>=40%)

In my analysis of the latest Vietnam Household Living Standard Survey Data (VHLSS), I found that the households who incurred impoverishment they had even smaller cut-off point of catastrophic (average of OOP/CTP

cutoff points/level of thresholds are various, as per the WHO, the health expenditure as catastrophic whenever it is greater than or equal to 40 per cent of capacity to pay.. for details go through this link-----

http://onlinelibrary.wiley.com/doi/10.1002/jid.1554/pdf

http://oii.igidr.ac.in:8080/jspui/bitstream/2275/105/1/WP-2010-001.pdf